Treatment strategies and outcomes
| Study (last name first author) | Tubbs | Carrington | Yousem | Akira | Travis | Ryu | Baloira | Kawabata | Individual cases | Total |
| Treatment | – | 26/40 (65) | – | – | – | 21/23 (91) | 12/12 (100) | 31/31 (100) | 61/67 (91) | 151/173 (87) |
| Removal of exposure | – | – | – | – | – | 4/27 (15) | 17/31 (55) | 6/61 (10) | 27/119 (23) | |
| Prednisolone | – | – | – | – | – | 21/23 (91) | 9/12 (75) | 31/31 (100) | 58/67 (86) | 119/133 (89) |
| Other first/second line treatment | – | – | – | – | – | – | 3/12 (25) | – | 12/67 (18) | 15/79 (19) |
| Follow-up years | – | 3.1±2.8 | 9 (0.8–22) | 3.2±1.3 | – | 3.1±3.7 | 1.9 (0.8–5.8) | 8.25 | – | – |
| Outcome | ||||||||||
| Improvement | 7/14 (50) | 16/26 (62) | 14/25 (56) | 2/6 (33) | – | 1/19 (5) | 4/8 (50) | 30/31 (97)+ | 37/66 (57) | 111/195 (57) |
| Stable | – | 3/26 (11) | 3/25 (12) | 3/6 (50) | 12/18 (67) | 12/19 (63) | 2/8 (25) | – | 5/66 (8) | 37/168 (22) |
| Worsened | 7/14 (50) | 7/26 (27) | 8/25 (32) | 1/6 (147) | 5/18 (28) | 1/19 (5) | 2/8 (25) | 5/14 (36) | – | 26/130 (20) |
| Death | – | – | – | – | 1/18 (6)# | 5/19 (25)¶ | – | 1/31 (3) | 14/66 (22) | 21/134 (16) |
| Transplantation | – | – | – | – | – | – | – | – | 2/66 (3) | 2/66 (3) |
| Relapse | – | – | – | – | – | – | – | – | 12/62 (18) | 12/62 (18) |
| PFT | ||||||||||
| Normalised | – | – | – | – | – | – | – | – | 7/39 (18) | 7/39 (18) |
| Improved | – | – | – | – | – | – | 6/9 (67) | – | 15/39 (38) | 16/48 (33) |
| Stabilised | – | – | – | – | – | – | 1/9 (11) | – | 10/39 (26) | 11/48 (23) |
| Deteriorated | – | – | – | – | – | – | 2/9 (22) | – | 7/39 (18) | 9/48 (19) |
Data are presented as n/n (%), mean, mean±sd, or mean (range). Available data from case series and individual cases are presented. Totals data are summarised as percentages of available data. PFT: pulmonary function test; CT: computed tomography. #: patient committed suicide; ¶: two patients died of cancer; +: five out of 14 patients with follow-up CT developed honeycombing during follow-up. Outcomes were not uniformly good as suggested by these data.